Title :
Learning a robust Tonnetz-space transform for automatic chord recognition
Author :
Humphrey, Eric J. ; Cho, Taemin ; Bello, Juan P.
Author_Institution :
Music & Audio Res. Lab. (MARL), New York Univ., New York, NY, USA
Abstract :
Temporal pitch class profiles - commonly referred to as a chromagrams - are the de facto standard signal representation for content-based methods of musical harmonic analysis, despite exhibiting a set of practical difficulties. Here, we present a novel, data-driven approach to learning a robust function that projects audio data into Tonnetz-space, a geometric representation of equal-tempered pitch intervals grounded in music theory. We apply this representation to automatic chord recognition and show that our approach out-performs the classification accuracy of previous chroma representations, while providing a mid-level feature space that circumvents challenges inherent to chroma.
Keywords :
learning (artificial intelligence); signal classification; signal representation; transforms; audio data; automatic chord recognition; chroma representations; chromagrams; classification accuracy; content-based methods; data-driven approach; de facto standard signal representation; equal-tempered pitch intervals; geometric representation; mid-level feature space; music theory; musical harmonic analysis; robust Tonnetz-space transform; temporal pitch class profiles; Accuracy; Harmonic analysis; Kernel; Time frequency analysis; Training; Transforms; USA Councils; Chord Recognition; Convolutional Neural Networks; Deep Learning; Tonnetz;
Conference_Titel :
Acoustics, Speech and Signal Processing (ICASSP), 2012 IEEE International Conference on
Conference_Location :
Kyoto
Print_ISBN :
978-1-4673-0045-2
Electronic_ISBN :
1520-6149
DOI :
10.1109/ICASSP.2012.6287914